Industrial information extraction through multi-phase classification using ontology for unstructured documents

•General purpose information extraction approach for industrial unstructured documents.•Novel method of multi-phase classification in two stages namely (i) sentence classification (ii) word classification.•Enhanced precision through iterative classification with feature weighing.•Unique normalization approach for transforming features with heterogeneous characteristics.•Extensible to non-industrial information extraction problems also on developing the domain ontology.